the ideas and inﬂ uence of alan williams · 2020. 10. 20. · the ideas and inﬂ uence of alan...

The Ideas and Infl uence of Alan Williams

The Ideas and Infl uence of Alan WilliamsBE REASONABLE – DO IT MY WAY!

Edited by

ANNE MASONResearch Fellow, Centre for Health Economics

University of York

and

ADRIAN TOWSEDirector, Offi ce of Health Economics

Radcliffe PublishingOxford • New York

Radcliffe Publishing Ltd18 Marcham RoadAbingdonOxon OX14 1AAUnited Kingdom

www.radcliffe-oxford.com

Electronic catalogue and worldwide online ordering facility.

© 2008 Anne Mason and Adrian Towse

All rights reserved. No part of this publication may be reproduced, stored in a retrieval system or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording or otherwise without the prior permission of the copyright owner.

Anne Mason and Adrian Towse have asserted their rights under the Copyright, Designs and Patents Act, 1998, to be identifi ed as Editors of this Work.

Neither the publisher nor the authors accept liability for any injury or damage arising from this publication.

British Library Cataloguing in Publication Data

A catalogue record for this book is available from the British Library.

ISBN-13: 978 184619 231 9

Typeset by Pindar New Zealand (Egan Reid), Auckland, New ZealandPrinted and bound by TJI Digital, Padstow, Cornwall, UK

Contents

Preface vii

List of contributors ix

1 A tribute to our friend and colleague, Alan Williams 1 Richard and Peggy Musgrave

2 Citizens, consumers and clients: Alan Williams and the political economy of cost–benefi t analysis 5

Robert Sugden

3 Alan Williams and cost–benefi t analysis in health care: comments on the paper by Robert Sugden 19

Bengt Jönsson

4 The public–private challenge in health care 27 Alan Maynard

5 Discussion of Alan Maynard’s paper: ‘The public–private challenge in health care’ 47

Dominique Polton

6 Resource allocation in health care: Alan Williams’ decision maker, the authority and Pareto 57

Anthony J Culyer

7 Discussion of Anthony Culyer’s paper: ‘Resource allocation in health care: Alan Williams’ decision maker, the authority and Pareto’ 75

Adrian Towse

8 Being reasonable about equity and fairness: looking back and extending the Williams way 85

Aki Tsuchiya and Paul Dolan

9 Equity and fairness in health and health care: looking up and extending ‘My way’ – a comment on Aki Tsuchiya and Paul Dolan’s paper 103

Jan Abel Olsen

10 Putting the ‘Q’ in QALYs 111 Paul Kind

11 Discussion of Paul Kind’s paper: ‘Putting the “Q” in QALYs’ 127 Ben van Hout

12 The measurement and valuation of public safety 133 Paul Dolan and Aki Tsuchiya

13 Discussion of Paul Dolan and Aki Tsuchiya’s paper: ‘The measurement and valuation of public safety’ 147

Martin Buxton

Index 153

vii

PrefaceBe reasonable : do it my way!

The sign on Alan Williams’ desk revealed his sense of humour, a man who invited and relished debate, but always recognising that intellectual pursuits were a means to a practical end.

Alan was a man of principles: as Bob Sudgen notes, Alan was not interested in ‘cookbook’ economics, but in developing guiding principles that embraced and encouraged active intellectual engagement and development. Many of the authors of the papers contained within this book testify to their encounters with Alan – their intellectual journeys: Bob Sugden recalls his student days and attributes his chosen career path to Alan’s infl uence; Ben van Hout reminisces about the impact Alan had upon Dutch efforts to develop quality of life measures; Bengt Jönsson echoes these sentiments, remembering how Alan’s courageous intellectual battles within Europe paved the way for younger academics; and Alan Maynard reminds us of the debate that raged (and perhaps rages still) between egalitarians and libertarians, both parties guilty, Williams observed, of comparing the ideal characteristics of their own ideology with the actual characteristics of the opposing ideology. As intellectual journeys criss-crossed, new ideas were born and principles were refi ned or revised; for Alan, principles were never meant to be followed slavishly or unthinkingly.

However, Alan was also a practical man: intellectual debates were not an end in themselves, but were for the purpose of tackling real-world issues. He wanted to help decision makers engage with the issues facing them; the role of the health economist, as he saw it, was to provide a clear framework through which important factors informing the decision-making process were made accessible and transparent. Thus Alan continued, well into his retirement, to propound applications for his work: always an egalitarian, he argued that there were equity grounds for discriminating against older people who had had their ‘fair innings’ and that these grounds had important implications for the way in which scare resources were allocated within tightly squeezed healthcare budgets. In recent years, Alan worked with Aki Tsuchiya on broader issues of

viii PREFACE

equity and fairness, exploring the implications of discriminating on the basis of gender or of socio-economic status.

Alan will perhaps best be remembered for his work within cost–benefi t analysis. The quality-adjusted life year (QALY), born of his desire to fi nd a generic outcome measure that would enable an assessment of the opportunity costs of healthcare interventions, synthesised the principled and the practical: life years added by a health intervention, adjusted for the quality of that life – not an end in itself but a means to achieving equitable health outcomes for all in the real world of limited resources. The role of the QALY as a tool for decision makers, enabling them to break out of the artifi cial constraints of ‘welfarist’ or ‘Paretian’ approaches, is discussed by Tony Culyer who sets out the key dividing points between Paretian and decision-making approaches to the application of economics in the allocation of resources in health care.

Within the healthcare sector, debates over the derivation and application of the QALY continue. Paul Kind reminds us that the quality element (‘Q’) of the QALY is critical: whose values should be used, how should values be combined and who should decide these issues? Kind argues that scope remains for methodological development, and that failure to address this has serious practical implications. Bob Sugden echoes these sentiments, questioning whether decision makers should decide, on behalf of the community, what the collective objective should be. Alan Maynard focuses on failure of healthcare systems to measure treatment effects, and advocates a system-wide application of patient-related outcomes measures (PROMs) as a way forward.

Alan’s work on the QALYs also inspired methodological work on outcome measurement in other fi elds. Paul Dolan and Aki Tsuchiya’s work on quality of life measurement in crime is in its early stages of development, with the SALY (safety-adjusted life year) proposed as a tool for measuring public safety. Their discussant, Martin Buxton, discusses the potential for a ‘super-QALY’ to embrace outcome measurement across different parts of the public sector, or even across multiple sectors.

Alan vigorously contested charges that cost–benefi t analysis was a ‘pseudo-science’. However, modern economic evaluation is not immune to the same allegations. Peggy and Richard Musgrave, who open this book with a tribute to Alan, refl ect on his concern with the philosophical and ethical issues that underpin decision making, issues that must still be faced. To recognise and appreciate Alan’s legacy, the task falls to the health economics community to ensure that we do not shirk our responsibilities: the need to be intellectually rigorous without being rigid; to keep in sight the practical implications of our work; to acknowledge the shortcomings within our discipline; and to move forward in the spirit of the Williams’ way.

Anne Mason, Research Fellow, University of YorkAdrian Towse, Director, Offi ce of Health Economics, London

September 2007

ix

List of contributors

Richard A MusgraveHH Burbank Professor of Economics, EmeritusHarvard University

Richard Abel Musgrave died on 15 January 2007 in Santa Cruz, California at the age of 96. A leading fi scal economist, he played a central role in shaping the modern fi eld of public fi nance.

Peggy B MusgraveProfessor of Economics, EmeritaUniversity of California, Santa Cruz

Robert SugdenProfessor of EconomicsUniversity of East Anglia, UK

Bengt JönssonProfessor of Health EconomicsStockholm School of Economics, Sweden

Alan MaynardProfessor of Health EconomicsUniversity of York, UK

Dominique PoltonDirectrice de la Stratégie des Études et des StatistiquesCaisse Nationale de l’Assurance Maladie, France

x LIST OF CONTRIBUTORS

Anthony J CulyerProfessor of Economics, University of York, UKSenior Scientist, Institute for Work & Health, Canada

Adrian TowseProfessor of Economics and DirectorOffi ce of Health Economics, London, UK

Aki TsuchiyaReader in Economics and Health EconomicsUniversity of Sheffi eld, UK

Paul DolanProfessor of EconomicsImperial College London, UK

Jan Abel OlsenProfessor of Health EconomicsUniversity of Tromsø and University of Oslo, Norway

Paul KindProfessor of Health EconomicsUniversity of York, UK

Ben van HoutProfessor of Medical Technology Assessment, University of UtrechtThe Netherlands and Scientifi c Director, Pharmerit

Martin BuxtonProfessor of Health Economics and Director of the Health Economics Research GroupBrunel University, UK

1

CHAPTER 1

A tribute to our friend and colleague, Alan Williams

. . . Richard and Peggy Musgrave

We take the occasion of Alan’s Memorial Conference to add our own words of esteem. Over the decades we travelled similar paths, both as colleagues and as friends. As colleagues, we shared fascination with the mysteries of ‘public fi nance’ and its powers to improve the world around us. As friends, we shared many visits, conferences and associations, crossing the Atlantic from east and west, from Vermont in the US to York in the UK. Hiking along the Pacifi c and in Vermont’s green hills, we recall Alan’s buoyant spirit, his gentle sense of humour spiced with sharp wit, a view of the world where not all was for the best, but neither was it beyond repair. Alan, striding fi rmly ahead, followed by June, slightly amused and with a twinkle of her own.

Alan’s earlier work, as did ours, focused on the basics of fi scal theory, the nature of public goods, the intractable issues of equity and the ways of implanting fi scal functions into the various forms of governmental organisation, central and local, unitary and federal. Later on, focus shifted to policy issues and it is here that Alan’s choice of health economics would lead him to his outstanding contributions, containing thoughts and directions which now enrich the papers of this conference.

Alan’s turn to health economics was indeed a brilliant move, guided by his knack for combining analytical insights with a sense for practical application; to clear the ground fi rst, before proceeding to more complex aspects. The purpose of health policy and research, as he saw it, is to improve the state of health of particular individuals, of groups and of nations, an essential ingredient of human well-being. Evaluation and comparison require measurement, a key issue to which Alan made one of his major contributions. The state of health as a policy concern has to be measured over time, traditionally done with reference to life expectation. Beginning with life expectancy as the point of departure, Alan then refi ned the concept, adding a human touch by allowing

2 THE IDEAS AND INFLUENCE OF ALAN WILLIAMS

for the quality of life that was experienced. Building on this framework, the quantitative analysis and assessment of policy outcomes could be advanced.

Through all this, just what is the nature of health and how does it compare with other benefi ts to be gained by economic activity? To begin with, is health improvement to be viewed as a private good, best left to the market, or is it a public good, calling for public intervention to overcome market failure? The answer depends on the particulars of the case and on how the problem is viewed. The individual patient, waiting for his pain pill to arrive, rightly views it as a private good, excludable and rival in consumption. Hence effi -ciency can be obtained in the market, and although government may enter to assure availability of medication as a matter of equity in distribution, it is not needed on effi ciency grounds. The answer differs, however, when con-sider ing medication given to prevent the spread of communicable disease. With consumption now non-rival, the market fails in its provision and public fi nancing is required.

Non-rival consumption may or may not be combined with excludability, and the role of the public hand differs accordingly. The advance of medical science, of medical ‘knowing’ of how to prevent and heal disease, is non-rival in use. Benefi ts to the consumer will be the larger, the more widely the given stock of knowledge is available, calling for its provision free of charge by the originator. This, however, would leave no compensation for those engaged in the advance of knowledge, and government must step in. Patent law might be used to preserve that incentive, limiting patent length so as to balance that gain against the immediate loss of not fully utilising the available stock of knowledge. Concern with the nature of public, private, and mixed goods thus provides an inseparable bond between health economics and public fi nance.

A further link between the two fi elds is provided by a shared concern with equity in distribution. In designing its budgets, the NHS applies two standards, choosing projects which are most effective in fi ghting particular diseases, and providing them in the most cost-effi cient way. Alan agreed, but asserted that effectiveness and effi ciency are not everything for ‘behind them, there lurks equity’. Who should pay and who should benefi t? How can health be measured, so that comparisons can be made and an equitable system be defi ned?

Fiscal equity, traditionally, has focused on the distribution of income across individuals, independent of its sources and uses. Severe inequality has been considered undesirable, with progressive taxation one of the correctives. But equity may also be viewed in ‘categorical terms’, and applied to the availability of goods which society views to be of particular importance (or ‘merit goods’ as they have come to be known). Among them, the availability of health services has been prominent. Publicly provided health services or support of privately provided services are used to correct for inequality in the patient’s capacity to secure provision. Once more, there exists a link between fi scal and health economics.

A TRIBUTE TO OUR FRIEND AND COLLEAGUE, ALAN WILLIAMS 3

Health services, like other goods, should be produced effi ciently, and resources be spread across various diseases so as to maximise benefi ts at the margin; and personal care should be given to reduce the burden of illness. All this matters, but just who should benefi t and who should pay? Equity enters through out the health problem, across both its healing and care functions, but what is most needed is a formulation of the problem in manageable form. The ethical and philosophical issues, such as entitlement to equal health, the right to ‘fair innings’ and the right to live or die, must also be faced at the end, and this is where Alan’s later work pointed. But the basics of measur ing health, essen tial for the guidance of health policy, must be addressed fi rst. Alan’s choice to pursue health economics was indeed a perfect match for his humanity, good sense and talents, which have been the key to his path-breaking work.

5

CHAPTER 2

Citizens, consumers and clients: Alan Williams and the political economy of cost–benefi t analysis

. . . Robert Sugden

INTRODUCTIONThis paper focuses on the work of Alan Williams, the cost–benefi t analyst, the predecessor of Alan Williams the leading health economist. In particular, it looks at Alan’s involvement in early debates about the proper role of cost–benefi t analysis (CBA) in the political process, about the relationships between the analyst, the agency commissioning a CBA and the wider community, and about the role of willingness-to-pay valuations in CBA. The issues that were involved in these debates remain relevant to current controversies about setting priorities in the delivery of health care – controversies that Alan continued to be engaged with until shortly before he died. The last time I met Alan – at a meeting in York in 2003 to discuss the methodology for a research project on ‘the social value of a QALY’ (quality-adjusted life year) – he was passionately defending a concept of ‘social value’ which can be traced back to the position he took in those early debates about CBA.

In trying to elucidate Alan’s position, I will use two main texts. (I am going back to an era when academic economists published much less than they do now.) The fi rst is the paper ‘Cost–benefi t analysis: bastard science? and/or insidious poison in the body politick?’, which Alan wrote in 1970, and which was published in the fi rst volume of Journal of Public Economics in 1972. The second is the book The Principles of Practical Cost–benefi t Analysis which he and I wrote together over far too much of the 1970s, and which fi nally appeared in 1978 (Sugden and Williams, 1978). I will do my best to reconstruct from a fading memory the discussions through which we arrived at agreed positions, and in this way to separate Alan’s views from mine.


COST–BENEFIT ANALYSIS AS MANAGEMENT CONSULTANCY: THE ‘BASTARD SCIENCE?’ PAPERIn striking contrast to the staid contents of more recent issues of Journal of Public Economics, Alan’s paper (which from now on I will refer to as ‘Bastard science?’) begins with a satirical verse written in the style of WS Gilbert, entitled The Management Consultant’s Lament. Alan reports that this verse was composed by a civil servant, JM Ross, as a response to Alan’s ‘own early activi-ties’ in the ‘booming business of management consultancy’ (Williams, 1972, p. 199). As this starting point suggests, Alan takes it as given that CBA is a form of management consultancy. His aim in the paper is to identify the features which distinguish CBA from other forms of consultancy, and to defend CBA against two charges: that it is a form of pseudoscience, and that, by sub sti-tuting economic analysis for political deliberation and negotiation, it usurps democratic processes.

Alan starts from the premise that the objective of CBA is ‘to assist choice[,] not to make choice’. He says that CBA is ‘a natural and logical extension’ of other tech niques which serve the same purpose, such as systems analysis, opera tions research (OR) and cost-effectiveness analysis (Williams, 1972, pp. 200–201). Viewed in this perspective, CBA has two distinguishing features.

First, CBA is ‘prescriptive’. Alan explains that, by this, he means that the analyst plays a ‘consultative role’; the analysis is ‘designed to help a client improve his situation’ (p. 202). It is signifi cant that the role of the analyst is defi ned in relation to that of a client: the client, presumably, is some agency (in the context of CBA, an agency of government) which is seeking assistance in making choices. The analyst’s prescriptions take the form of advice about how the client agency can best achieve its own objective. This does not mean that the analyst’s attitude towards the client has to be completely uncritical. Alan points out that clients often lack a clear sense of the nature of ‘the problem’ that they are asking the analyst to solve: ‘any OR or CBA practitioner who accepts the client’s initial formulation of the problem uncritically is heading for disaster’ (Williams, 1972, p. 204). Still, it is clear that the analyst’s role is to help the client to conceptualise its problem, as it perceives it, and not to prescribe what the client should be aiming to achieve.

The second, and most distinctive, feature of CBA is that it ‘[tries] hard to render inputs and outputs commensurable’. In its ideal form – an ideal that is not currently attainable – ‘CBA will have all inputs and outputs evaluated in money terms’ (Williams, 1972, pp. 200–202). However, Alan is at his most forceful in insisting that this does not mean that valuations should be based on individual willingness-to-pay.

Using Aaron Wildavsky as a spokesman for the claim that CBA is an insidi ous poison in the body politick, Alan quotes the following comment on CBA:

The economic model on which cost–benefi t analysis depends for its validity is based on a political theory. The idea is that in a free society the economy is to

CITIZENS, CONSUMERS AND CLIENTS 7

serve the individual’s consistent preferences revealed and rationally pursued in the market place. Governments are not supposed to dictate preferences nor make decisions.

(Wildavsky, 1966, p. 297)

Alan sees this as a gross misrepresentation: ‘This is so patently false that if I had not had an opportunity to read and re-read this statement carefully I would not have believed it possible that such a well-informed and normally judicious observer could have made it’ (Williams, 1972, p. 214). Alan accepts that ‘much of the analysis that goes into most cost–benefi t studies is based on the assumption that the market does constitute a good guide to the value of costs and benefi ts’; but he claims that the practice of CBA allows this assumption to be relaxed. Some of the relaxations he describes seem to me to be in the spirit of Wildavsky’s account of CBA. For example, Alan says that CBA ‘may require . . . valuations inferred from market behaviour to be placed on items which have no market values’; but this is a case in which the government is eliciting preferences, not dictating them. Alan’s crucial objection is that CBA ‘may require . . . the use of valuations postulated by the policy-makers or decision-makers themselves’ (Williams, 1972, p. 214).

The legitimacy of such postulated values is a recurring theme in the paper. For example, he accepts that if ‘a client (say a Minister of Transport) [says] that for the purposes of this analysis I want variable X valued at £Y per unit (where X might be the intrinsic value of life)’, then ‘£Y represents a precise statement about the client’s policy trade-off between X and the other variables in the analysis’, and so is a legitimate datum for CBA. This kind of postulated value is ‘arbitrary’ in the sense that it is not supported by evidence, but it is not ‘irresponsible’ because ‘it does imply a commitment to allocate resources in a particular way for which the client will have to answer’, and ‘it provides a precise focus for discussion by others (say Cabinet colleagues) who might wish to infl uence his judgements on this matter or modify their own’ (Williams, 1972, p. 210). Notice that, in this account, postulated values are presented as internal to the government machine. There is no suggestion of public debate about these values, apart from the hint that the government may ultimately have to answer to its electorate.

Here is how Alan sums up his position on the status of market-based and postulated values in CBA:

I am doing no more than accepting the well-recognised proposition that you cannot ascribe values without making value judgements. Market prices are acceptable if the value-premises underlying market behaviour are acceptable (e.g. accepting the prevailing distribution of economic power and the relevance of individual market-oriented valuations to the making of social judgements) . . . Postulated prices are acceptable if it is believed that the value-premises


underlying both market and imputed prices are misconceived for the purpose in hand (e.g. if one accepts the propriety of a paternalistic or collectivist basis for valuation).

(Williams, 1972, p. 220)

The reader is entitled to ask: Who is making the value judgements here? To whom are postulated prices acceptable or unacceptable? The logic of the paper as a whole seems to allow only one answer: the client.

Summing up: Alan’s position in ‘Bastard science?’ is that CBA is a par ticu-larly ambitious form of management consultancy. It is scientifi c in the sense of being systematic: ‘assumptions are required to be explicit, evidence presented, results communicable and replicable’ (Williams, 1972, p. 203). It does not usurp the political process because it does not claim to make decisions and does not impose any substantive value judgements; it merely assists clients to formulate their own objectives and to pursue them effectively.

TWO OPPOSING VIEWS OF COST–BENEFIT ANALYSIS: THE PRINCIPLES OF PRACTICAL COST–BENEFIT ANALYSISBefore discussing the text of The Principles of Practical Cost–benefi t Analysis (Sugden and Williams, 1978), I need to explain how the book came to be written, and Alan’s and my respective roles as co-authors.

My becoming an economist was very largely due to Alan. I was an under-graduate at York between 1967 and 1970, studying history and economics. The topic in economics that most interested me was CBA, which I learned about from Alan’s lecture course on investment appraisal. For my fi nal-year project, I chose to do a retrospective cost–benefi t study of the recent closure of the Whitby–Scarborough railway. Alan was my supervisor, and went far beyond the requirements of duty in helping me to make contact with various agencies that were able to provide me with data I needed. He had gently suggested to me that the plan of carrying out a complete cost–benefi t study was over-ambitious for one student over a summer vacation, but I wanted to (and, I think, did) prove him wrong. While reading the reports of previous cost–benefi t studies of rail closures, I had my fi rst genuinely original economic idea: I realised that the methodology that had been used by the Ministry of Transport in its most recent rail-closure study made a theoretical error in its treatment of changes in fare revenue. When I told Alan that I had discovered this error, I sensed that he was sceptical, but he listened to my argument in his characteristically open-minded way. After some thought, he agreed that I was right. He encouraged me to write a paper on the subject and send it to the Bulletin of Economic Research; this became my fi rst publication (Sugden, 1972). The following summer, I had my fi rst paid job as an economist, working for Alan on a project he had persuaded the University of York to fi nance, investigating whether


the apparent scarcity of private-sector student accommodation in the city could be reduced by making suitable adjustments to bus routes. (This is a good illustration of Alan’s conception of the creative role of the analyst. The university perceived its problem as a shortage of student accommodation in areas accessible to the campus; Alan saw that the problem could be restated as the poor accessibility of otherwise suitable areas).

After a year studying for a master’s degree in Cardiff, I returned to York in 1971, as by far the economics department’s youngest and least-qualifi ed lecturer. Shortly after this, Oxford University Press invited Alan to write a text book on CBA. He had a very clear idea about the kind of book that was needed – an idea expressed in the apparently paradoxical title he chose for it. It should be practical in orientation, teaching would-be analysts how to provide useful advice to clients, using simplifi ed case studies as a method of instruction and as training exercises. But it should not be a cookbook, presenting CBA as a set of techniques to be applied mechanically; it should induct its readers into the general principles on which CBA was based. In fact, it should follow the strategy of his investment appraisal lecture course. Since he didn’t have time to write this much-needed book on his own, he suggested that we took on the project as co-authors – I doing most of the actual writing, but under his general guidance and, of course, with his reputation to sell the book. (I should say that he insisted that almost all the royalties came to me.)

In many ways, The Principles of Practical Cost–benefi t Analysis was, for me, the equivalent of writing a PhD thesis (something I never did); Alan’s role was similar to that of a thesis supervisor, except that the fi nal text had to be in a form that he could sign up to. At fi rst, I didn’t expect this to be much of a problem. However, just as PhD supervisees usually do, I gradually developed a perspective of my own, and it became more diffi cult for us to reach agreement on what our book should say. Increasingly, I felt the pull of currents of thought in economics which ran counter to Alan’s understanding of CBA.

In the early 1970s, economists’ views about the nature of CBA began to polarise. One school of thought was essentially that of ‘Bastard science?’. On this view, CBA is a branch of economic planning; the key ideas it imports from economics are those of optimisation theory. There are no a priori restrictions on the objective function to be maximised. This position was taken by Ian Little and James Mirrlees in their infl uential text, Project Appraisal and Planning for Developing Countries (1974). However, there was an opposing school of thought, represented by Ezra Mishan’s equally infl uential Cost–benefi t Analysis (1971). On this view, CBA is the applied branch of what in 1971 was still ‘new’ welfare economics – that is, the welfare economics of the Kaldor–Hicks compensation test. The key ideas that CBA imports from economics are those of welfare economics, consumer theory and price theory, particularly the concepts of producers’ and consumers’ surplus. The task of the cost–benefi t analyst is to investigate whether the total amount that those who would gain from a policy proposal would be willing to pay for their gains is greater or less than the


total amount that losers would be willing to accept as compensation. On this account, the cost–benefi t analyst is an independent specialist, committed to applying a particular set of evaluative principles. A client who employs such a specialist cannot ask for those principles to be changed: he pays to be told what they imply in a specifi c case. There is simply no place in CBA for postulated values. Mishan, I think, would have endorsed the quotation from Wildavsky (see pp. 6–7) which Alan thought patently false.

Another pull came from the theory of public choice, one of the intellectual growth areas of the 1970s and part of the wave of ‘New Right’ thinking which paved the way for Margaret Thatcher’s election victory in 1979. A central theme of public choice theory was the implausibility of (what was then) the conventional economist’s view of government as a benevolent despot or deus ex machina, a neutral force which could be called in to correct market failures. The public choice literature taught economists to see political failure as just as much a problem as market failure, and to model the behaviour of government agencies in terms of the interactions of individually motivated actors. Alan’s conception of the ‘client’ for CBA began to sound suspiciously like the non-existent benevolent despot.

Also built in to much of the theory of public choice was the hypothesis that when individuals act as voters and when they act in the market, they are acting on the same preferences: at the level of the individual, there is no distinction between the ‘citizen’ and the ‘consumer’. (Let me say in passing that I am no longer persuaded of the truth of this hypothesis.) Anyone who accepts this hypoth esis will have diffi culty with Alan’s idea that one might conclude that ‘the value-premises underlying both market and imputed prices are mis con-ceived for the purpose in hand’ and that one might instead accept ‘the propriety of a paternalistic or collectivist basis for valuation’ (Williams, 1972, p. 220). The value-premises underlying market prices, one would have to conclude, are simply the preferences of the individuals who act in the market. As voters, those same individuals have the same preferences. So when a government agency declares that these value-premises are misconceived, it is declaring that it is choosing not to act in accordance with the preferences of the people who elect it and pay for it. How can this be justifi ed to those people?

In responding to these currents of thought, I was infl uenced by other York colleagues, particularly Mike Jones-Lee and Tony Culyer, and later by James Buchanan. Mike Jones-Lee was a fi rm adherent of Mishan’s approach to CBA. His pioneering work on eliciting ordinary individuals’ valuations of reduc-tions in risks of premature death was showing that it was possible to construct willingness-to-pay valuations even of ‘statistical life’ (Jones-Lee, 1976); he convinced me that this was a better way for CBA to deal with risks of death than the use of postulated values. More generally, the analysis of individual willingness to pay was in the spirit of the York department’s tradition of applied microeconomics. At the time, Tony Culyer was producing imaginative analyses of real-world problems using only the simple components of Marshallian price


theory, with a touch of populist scepticism about governments which increas-ingly appealed to me. I had the good fortune to be able to spend the summer of 1977 at Buchanan’s Public Choice Centre in Blacksburg, Virginia, and felt an immediate attraction to his contractarian form of normative economics, in which the role of government is not to judge what is good for society, but to implement projects which individuals recognise as being in their mutual inter-est. This was just after Alan and I had fi nished our book, but the ideas which crystallised for me in Blacksburg had been forming long before that.

So how did Alan and I manage to complete our book? Despite the famous placard on his desk, Alan’s line was never ‘Be reasonable: do it my way’. Rather, he looked for ways in which we could acknowledge our fundamental differences within the structure of a practically oriented textbook. The original plan of the book, as designed by Alan, served us well. As in ‘Bastard science?’, we conceptualised the appraisal process as a dialogue between ‘the analyst’ and ‘the decision maker’ (Alan’s ‘client’). In order for analysis to be possible, we said, the decision maker’s objective needed to be specifi ed. The kind of analysis that was required depended on the nature of that objective. In many contexts, even in the public sector, agencies are expected to act on ‘commercial criteria’ – that is, to use fi nancial appraisal. So we began by explaining the principles of fi nancial appraisal. This relatively uncontroversial material took up the fi rst third of the book.

We then suggested that public decision makers might be expected to take account of a wider range of policy effects than can be encompassed by fi nan-cial appraisal. As one way of doing this, we introduced the compensation test. At this point, we distinguished between two ways of thinking about CBA – the decision-making approach (Alan’s preferred approach, in which the objective is whatever the decision maker chooses it to be) and the Paretian approach (Mishan’s approach, in which CBA is seen as an application of Paretian welfare economics, and the objective is given by the compensation test). However, we began by playing down this distinction, presenting the com pen sa tion criterion as a simple and tractable way of taking account of a wide range of effects in a consistent way; as such, it provides a credible ‘fi rst approximation to a full statement of government objectives’, even within the decision-making approach (Sugden and Williams, 1978, p. 93). This kept us going for another third of the book, in which we dealt with shadow pricing, consumers’ and producers’ surplus, imputed prices and uncertainty. It was only in the fi nal third of the book that we introduced the idea of postulated values – for particular goods such as health care, for weighting gains and losses to different income groups, and in the form of a postulated social discount rate. Here, things became more awkward as we repeatedly reminded the reader that the legitimacy of postulated values was a controversial issue, and tried to treat the two approaches to CBA even-handedly.

In the fi nal chapter, entitled ‘Epilogue: the analyst, the decision-maker and the community’, we reworked some of the themes of ‘Bastard science?’, with


the signifi cant difference that the analyst and decision maker were placed in a triangular relationship with ‘the community as a whole’. We declared that ‘the public decision-maker is entrusted to act on behalf of this community’ (Sugden and Williams, 1978, pp. 220–30). Clearly, the intention was to qualify Alan’s earlier account of CBA as management consultancy, by making both analyst and client responsible to a wider community. My recollection is that the idea that we should do this was originally mine, but that Alan worked over my drafts quite closely, suggesting revisions and making sure that he could endorse the sentiments expressed.

The idea that the analyst has responsibilities beyond those of a consultant to a client was developed in two ways. The fi rst (which, as far as I can recall, Alan supported wholeheartedly), appeals to a professional ethic of intellectual honesty. In the perspective of the decision-making approach, one of the principal virtues of CBA is consistency. It is by making different decisions of the same agency consistent with one another, and consistent with a common objective, that CBA promotes effi ciency in the allocation of resources. However, if the benefi ts of effi ciency are to be achieved, the CBA must be done in good faith and not as mere window-dressing. The more freedom decision makers have to specify their own objectives and to postulate their own values, the more scope there is for self-serving window-dressing. I think I felt at the time – and I certainly believe now – that this is a serious weakness of the decision-making approach. The less discretion there is for CBA methodology to be tailored to specifi c appraisals, or even to the interests of specifi c commissioning agencies or governments, the more confidence one can have that the results are genuinely informative. On these grounds, I now favour the convention that CBA should be based on individual willingness to pay (see Sugden, 2005). But back in the mid-1970s, I was still making up my mind. Alan and I were able to sign up to the claim that, if consistency is a virtue of CBA, the analyst’s professional role must include the advocacy of consistency, even in cases in which the client would prefer to be inconsistent. Thus, in reporting the results of a cost–benefi t study, the analyst has a professional responsibility to point out the wider implications of using whatever values have been postulated for that study (Sugden and Williams, 1978, pp. 231–6).

The second form of responsibility is to the community as a whole. My recol lec tion is that we had more diffi culty agreeing on the passages which deal with this issue. We discussed it in relation to another virtue of CBA, explicitness. We pointed out that the explicitness of CBA is not always desired by clients: their inter ests may sometimes be better served by obfuscation. However (and here the infl uence of public choice theory can be detected), explicitness might have a cor respond ing value to the community as a whole, as a mechanism of accountability:

Thus cost–benefi t analysis should not be seen solely as a service to the decision-maker. If the analysis is not secret, it exposes the logic of the decision-maker’s


actions to the scrutiny of those to whom he is accountable . . . [It] carries a stage further the function of traditional fi nancial accounting. The obligation on the part of privately owned fi rms and public agencies to keep fi nancial accounts is a very effective deterrent against embezzlement and fraud by managers, public offi cials and politicians. The obligation to justify public decisions within the framework of cost–benefi t analysis discourages a much subtler form of abuse of responsibility – that of taking decisions on behalf of others by using criteria that these others would not approve.

(Sugden and Williams, 1978, p. 240)

Signifi cantly, our concept of accountability allowed political decision makers to choose objectives and postulate values. We offered the following understanding of the political process (cautiously prefaced by ‘one can argue that’):

the role of the analyst is to assist, not simply a decision-maker, but a decision-making process that has the assent of the community as a whole . . . The decision-maker is responsible for making a decision, according to his own lights, but he is responsible to the community. His right to decide stems from the consent of the community, expressed through the political system. The community, then, ought to have the right to call upon the decision-maker to account for his decisions.

(Sugden and Williams, 1978, p. 241)

This formulation leaves room for CBA to be based on the kind of ‘collectivist’ value premises that featured in ‘Bastard science?’. The claim that these premises are social values is still only a value judgement made by the decision maker on behalf of the community: there is no requirement that this judgement is endorsed by the members of the community themselves. But there is a recognition that the decision maker has to account for his judgements to the people on whose behalf he claims to make them.

CITIZENS AND CONSUMERS IN ALAN WILLIAMS’ HEALTH ECONOMICSThe views expressed in the Epilogue of The Principles of Practical Cost–benefi t Analysis (Sugden and Williams, 1978) seem to occupy an intermediate position between the ‘Bastard science?’ account of CBA as management consultancy and the conception of the role of the health economist that is implicit in Alan’s later work.

On my (far from complete) reading of Alan’s work in health economics, his central research programme was a form of CBA, in the main senses in which he defi ned CBA in ‘Bastard science?’. It was prescriptive, designed to assist choices about the allocation of resources within publicly fi nanced


healthcare systems. And it was committed to the aim of making inputs and outputs as commensurable as possible. The whole point of the QALY, for which he was such a dedicated advocate, was to try to make the outputs of the healthcare industry commensurable with one another. Alan was always clear-sighted in recognising that priority setting was an inescapable part of medical decision making, and that allocative effi ciency implied a shadow-price for a QALY (or shadow-prices for QALYs accruing to different classes of benefi ciary). However, he never accepted that QALYs should be valued in terms of willingness to pay. (This was what was at dispute in the meeting in 2003, which I mentioned in the introduction. In a reprise of the York of the 1970s, but with some interesting re-alignments of ideological positions, Mike Jones-Lee and I argued for the willingness-to-pay approach, while Tony Culyer inclined to Alan’s position.) In this sense, Alan maintained his commitment to the legitimacy of postulated values.

However, Alan’s conception of the ‘client’ – the audience to which his work was addressed, and the source of postulated values – seems to have undergone some change. He sought to engage, not only with public decision makers, but also with the wider community of people who use public health services and who pay for them through their taxes. Crucially, he sought to engage with them as citizens and not as consumers.

In a paper of 1988, with the title ‘Priority setting in public and private health care: a guide through the ideological jungle’, Alan nails his own colours to the mast. He argues that, despite the messy mix of public and private elements that we fi nd in all real healthcare systems, there is a fundamental ideological divide between ‘libertarian’ and ‘egalitarian’ viewpoints about the provision of health care. In a libertarian healthcare system, the ‘dominant ethic’ is ‘willingness and ability to pay’, while in an egalitarian system, it is ‘equal opportunity of access for those in equal need’ (Williams, 1988, p. 174). Alan asks us to commit ourselves to one ethic or the other: ‘Each of us must decide for ourselves where we stand in that particular confi guration of attitudes, and be honest with ourselves and with others about it’. He declares himself an egalitarian. Although these commitments are made by us as individual citizens, Alan argues that what is ultimately required is a collective decision about which of the two ideological positions is to govern the provision of health care in a given political community.

He rejects the suggestion that there can be a mixed solution, of the kind that a contractarian might favour, in which there is a socialised system of health care for those who value social insurance and are prepared to pay for it collectively, and a market-based system for those who choose to opt out and provide for themselves:

I feel quite strongly egalitarian, and would aim to make the public system stronger and the private system weaker, in any community on which I depend for health care. But I also recognise the need, in a democratic country, to


respect the ideological position of the minority, provided it is not actually subversive. The trouble with private systems, in my view, is that they become ‘subversive’ if permitted to play a signifi cant role in a mixed system, because public systems rely on strong feelings of social solidarity (the rich must help the poor, the healthy the sick, the wise the foolish, the well-informed the ignorant, and so on), whereas private systems exist precisely to enable the rich, healthy, wise and well-informed to ‘opt out’ and look after themselves.

(Williams, 1988, p. 182)

The implication is that a fundamental collective decision has to be made about which set of values is to predominate. When it comes to the setting of priorities, each type of healthcare system ‘has to be judged according to its own lights, i.e. according to its own ideology’ (Williams, 1988, p. 183). In his work as a health economist, Alan starts from the premise that, in the UK, a collective decision has already been made in favour of the use of egalitarian values. This prior decision sets the ground rules for political debate about healthcare priorities.

Thus, in a discussion piece on age-based rationing, published in the British Medical Journal (BMJ) in 1997, Alan poses the question: Whose values should count in a social insurance setting? He asks us to suppose that older people are willing to pay more than younger people for health improvements for them-selves. Is that relevant for the setting of priorities in the NHS? Alan insists it is not:

But did we not take the NHS out of that [private market] context precisely because as citizens (rather than as consumers of health care) we were pursuing a rather different ideal – namely, that health care should be provided according to people’s needs, not according to what they were each willing and able to pay[?] A person’s needs (constituting claims on social resources) have to be arbitrated by a third party, whose unenviable task it is to weigh different needs (and different people’s needs) one against another. This is precisely what priority setting in health care is all about. So the values of the citizenry as a whole must override the values of a particular interest group within it.

(Williams, 1997a)

So, Alan is declaring, the setting of healthcare priorities in the NHS must be based on judgements about relative need, made by a ‘third party’. The reference to the ‘unenviable task’ strongly suggests that the third party is a political or professional decision maker – someone like the ‘client’ of ‘Bastard science?’ or the ‘decision-maker’ of Principles of Practical Cost–benefi t Analysis. The implication seems to be that priorities are set by the decision-maker, on behalf of ‘the citizenry as a whole’.


Much more than in his earlier work on CBA, however, Alan the health economist wants to draw the citizenry into the priority-setting process. A major part of Alan’s research programme in health economics has been con cerned with eliciting, from representative samples of the population, citizen-perspective judgements about marginal trade-offs between different healthcare benefi ts and different classes of benefi ciary. In the BMJ paper, he appeals to evidence from surveys which show that most people, including the old themselves, favour giving priority to the health care of the young (see also Williams, 1997b, p. 118). The implication is that this is relevant evidence for decision makers. In addition, Alan clearly wanted to foster public debate about priorities. Thus, in the BMJ paper, he offers the hope that ‘reasonable limits’ on the demands for health care that the old can make on their fellow citizens can be set ‘with fairly general consent’, and (as an elderly man) he appeals to the members of his generation to exercise restraint in the political demands they make on the healthcare system.

As an example of Alan’s conception of the relationship between the decision maker and the community, consider the following passage. He is responding to an argument by Amartya Sen that the principle of non-discrimination between the sexes in the delivery of health care should take priority over the pursuit of equality in lifetime health experience. (Since men have shorter life expectancy than women, equalising access to QALYs between the sexes would require discrimination in favour of men.) Alan notes that, following Sen, one might draw up a list of ‘possible axes of discrimination’ and make a priori moral judgements about the acceptability or unacceptability of discrimination on each axis:

But this is not the path I would take. I would prefer to fi nd out from people generally at what point they would be willing to discriminate (and how strongly) when told the amount of damage such a moral constraint is causing to the important aim of reducing inequalities in lifetime experience in health . . . I would then compare the median trade-off rate from my representative sample of the population with the shadow price of the restriction as it currently operates, and then decide, in the light of the results, whether to accept the status quo or start thinking of ways of tightening up or relaxing the moral constraint.

(Williams, 2003, p. 65)

I take it that, by ‘the path I would take’, Alan means the path that he would take as a public decision maker, charged with setting healthcare priorities. He is disagreeing with Sen on two levels. As a good economist, he is rejecting the idea of absolute moral constraints in favour of marginal trade-offs between different objectives. But he is also rejecting the idea that public decisions should be governed by the moral principles that decision makers (or moral philosophers) deem to be good or right. Instead, he proposes to investigate


the judgements that are in fact made by informed citizens. Notice that his survey respondents are not being asked to compare different combinations of healthcare benefi ts and costs for themselves; they are being asked to play the role of third-party arbitrator. Notice also that Alan’s decision maker appears to be reserving to himself the fi nal decision about what to do ‘in the light of the results’. It seems that the survey of citizen judgements is intended to inform that decision, not to make it.

There is, I suggest, an essential continuity from the account of CBA in ‘Bastard science?’, through the ‘decision-making approach’ of The Principles of Practical Cost–benefi t Analysis, to Alan’s approach to health economics. Throughout, Alan has been committed to a conception of political decision- making in which policy options are assessed in relation to an objective which in some way expresses a collective judgement of the relevant community. His inclination, I think, has always been to see applied microeconomics as addressed to an imagined public decision maker who is ultimately responsible for deciding, on behalf of the community, what the collective objective is to be. He has consistently rejected the idea that this objective can be constructed from the preferences that individuals reveal as consumers, whether in the market or in response to questions about their willingness to pay for private benefi ts. However, and particularly in his later work, he has expected public decision makers to be responsive to the judgements that individuals make as informed citizens. My contractarian leanings prevent me from endorsing this approach; but no one can do more than Alan to make it reasonable.

ACKNOWLEDGEMENTSMy work on this paper was supported by the Economic and Social Research Council (award no RES 051 27 0146).

REFERENCESJones-Lee M (1976) The Value of Life: an economic analysis. London: Martin Robertson

and Chicago University Press.Little I and Mirrlees J (1974) Project Appraisal and Planning for Developing Countries.

London: Heinemann Educational.Mishan E (1971) Cost–benefi t Analysis. London: Allen and Unwin.Sugden R (1972) Cost–benefi t analysis and the withdrawal of railway services. Bulletin

of Economic Research 24: 23–32.Sugden R (2005) Coping with preference anomalies in cost–benefi t analysis: a market-

simulation approach. Environmental and Resource Economics 32: 129–60.Sugden R and Williams A (1978) The Principles of Practical Cost–benefi t Analysis. New

York, Oxford: Oxford University Press.Wildavsky A (1966) The political economy of effi ciency: cost–benefi t analysis, systems

analysis and program budgeting. Public Administration Review 26: 292–310.


Williams A (1972) Cost–benefi t analysis: bastard science? and/or insidious poison in the body politick? Journal of Public Economics 1: 199–225.

Williams A (1988) Priority setting in public and private health care: a guide through the ideological jungle. Journal of Health Economics 7: 173–83.

Williams A (1997a) The rationing debate: rationing health care by age: the case for. British Medical Journal 314: 820.

Williams A (1997b) Intergenerational equity: an exploration of the ‘fair innings’ argument. Health Economics 6: 117–32.

Williams A (2003) Comment on Amartya Sen’s ‘Why health equity’. Health Economics 12: 65–6.

19

CHAPTER 3

Alan Williams and cost–benefi t analysis in health care: comments on the paper by Robert Sugden

. . . Bengt Jönsson

INTRODUCTIONFor me, Alan Williams the cost–benefi t analyst is synonymous with Alan Williams the leading health economist. His major research was on the methods and appli cation of economic evaluations to resource allocation in health care.

Bob Sugden (BS) rightly points out that Williams’ 1972 paper was inspired by the shift of interest in public economics from public fi nance to public expenditure analysis, and particularly PPB (planning, programme, budgeting) analysis, that occurred in the 1960s, and which drew critique from political scientists like Wildavsky (Wildavsky, 1973). Economists thus come into a role as ‘management consultants’ to governments, to help improve the use of the resources raised from taxation, thus competing with the political scientists as advisors. To meet the critic there was a need to defi ne the criteria for what was included, and not included, in a cost–benefi t analysis.

Alan Williams’ (hereafter known affectionately as ‘AW’) ‘Bastard science?’ paper triggered my fi rst visit to York. Although I approached him as a ‘cost–benefi t’ analyst, he introduced me to the concept of ‘health economics’ (Cooper and Culyer, 1973), and the visit prompted a shift in my research interest towards application of CBA in health. The idea for my planned PhD thesis was very close to Alan’s idea for the textbook that he was invited to write at about the same time: to review the published studies in the fi eld with the objective of systematising and comparing, and to fi nd a possible common theoretical base (principles). It should not be yet another ‘bad example’ of an applied study without a clear method, nor a cookbook of how to perform studies. The


aim was to reveal and develop the practices and principles involved. The work on my dissertation, which benefi ted very much from continued contacts with AW and members of the Health Economics Working Group (HEWG) during that time, was thus parallel to AW’s and BS’s work with the textbook to be published in 1978 (Sugden and Williams, 1978).

When the dissertation was completed and ready to be defended, the faculty chose AW to act as opponent (Jönsson, 1976). The defence of the dissertation is an important academic institution in Swedish academic life, where a leading researcher in the fi eld is asked by the faculty to review and critique the thesis in public. It is not so common that the opponent is asked to document his ‘opposition’, but AW’s comments were considered so ‘helpful and enlightening’ (quote from preface by Lars Söderström, Head, Department of Economics, Lund University) that he was asked to include his notes in the department’s series of seminar papers (Williams, 1976). Since his comments probably are not much known outside Lund, I will quote them extensively to contrast, or rather reinforce the observations BS makes in his paper. I will raise the following questions for discussion:1 what was AW’s position on CBA in general and in its application to health

in particular?2 how can it be explained, what was the rationale?3 what were the implications and consequences of his position?

ALAN WILLIAMS’ POSITION ON CBAI have no objections to BS’s description of AW’s position, which I would like to summarise as follows:

The role of CBA in the decision-making processShould be scientifi c, yet practical and based on explicit values ●Should be aimed at helping decision makers do better. ●

Critique of Paretian welfare economicsBased on willingness to pay (WTP) and ability to pay which is rejected as ●a resource allocation principle in health careThe decision-maker approach (management consultancy?) is an ●alternative.

The methodology of CBAMarket-based values and willingness to pay is only acceptable for ●valuation of resources (the cost side)Postulated values for health benefi ts based on decision-maker values ●Developed into cost per QALY as a general method for the application of ●CBA to health.

ALAN WILLIAMS AND COST–BENEFIT ANALYSIS IN HEALTH CARE 21

His position could be contrasted to the essential positions taken in my dis-sertation:

acceptance of a public healthcare system where everybody has access to ●health care, but where few, if any, decision makers care about effi ciency in resource allocationCBA is needed, particularly for decisions about new technology, but ●practice must be improved – studies are not comparablethere is ● government failure as well as market failure, and CBA is a method to correct bothwelfare economics should guide application and method for CBA ●the value of health improvements should be assessed through expressed ●WTP from a citizen perspective (insurance method).

Alan started his opposition with the statement: ‘I see my role as being that of bringing to Bengt Jönsson’s work the perspective of a non-Scandinavian observer of the cost–benefi t scene, who has recently devoted a large part of his time to the planning and evaluation of health services (especially in Britain) with the object of exploring the potential of the tools and insights of economics’. AW’s views on CBA were formed by his attempts to apply economics to the healthcare sector. For me, AW the CBA analyst and AW the health economist are very much the same thing. He had formed his position over a number of years, as evident from the four papers he is referring to in the notes for his opposition (Williams, 1974a–c, 1977). His views expressed in his comments on the thesis refl ect his position in the middle of the work on his book together with BS.

AW was very determined in the view that the role of CBA is to assist decisions, not to make them (the insidious poison critique). While I subscribed to the same view ‘The purpose of the thesis is to demonstrate “the way in which economic theory and method can be used to clarify the resource problems in the medical sector, to assist choice not to make choice”’(AW quoting from the dissertation). However, I have a feeling that he was not totally satisfi ed with the word ‘clarify’ in my statement. Even if the ‘management consultant’ is not the responsible decision maker, his role is to help the decision maker to arrive at a conclusion, given his objectives and the constraints. To limit the role of the economist to investigate whether a particular resource allocation was effi cient according to the Kaldor–Hicks criteria, which was the position I explored in the thesis, was not in his mind, as we will see later.

WHAT IS THE ECONOMIST’S PARTICULAR CONTRIBUTION?AW stressed the need for a clearly defi ned decision problem. The role of the health economist, in his role as management consultant, was to help the decision maker to defi ne the problem. He stressed that it is ‘hardly ever a decision about “ALL OR NOTHING”’ and he continued ‘of all the major


themes in the study, this is probably the most important contribution of economists to problem formulation (and why it is so silly to regard them simply as hirelings to be brought in after the real work has been done, to add a touch of cosmetic costing to an ill-designed study)’ (Williams, 1976).

Another role for the economist is to ‘make sure all relevant costs/benefi ts are included, but only once’. Thus AW subscribed to social cost perspective and not a specifi c decision-maker perspective. He noted that ‘CBA is not the same as fi nancial appraisal’, it should have a ‘GNP [gross national product] orientation (opportunity cost of time)’ and that also resources ‘not in GDP [gross domestic product] (time lost for persons not in the labour force)’ should be included (Williams, 1976).

In a later paper (Williams, 1981) he develops this, stating that ‘When we assume that all changes in real resources associated with the alternative under investigation have been measured, it is useful to classify them as:a. changes in resources used in service productionb. changes in resources used by patients and their helpersc. changes in gross domestic product’.

While he also stated that ‘the benefi t measure shall not depend on the wealth or “economic” value of the individual’, he never addressed the problem that benefi ts/costs should be included ‘only once’. It may, for example be diffi cult to measure health improvements as the number of QALYs gained, without differences in resources (costs and income) between states not affecting the estimate. See for example the discussion about to what extent indirect costs are included in measures of the utility of a health state (Gold et al., 1996).

THE PURPOSE OF COST–BENEFIT ANALYSIS, AND ALAN WILLIAMS’ CRITIQUE OF PARETIAN WELFARE ECONOMICSAW subscribed to the view that the role of CBA was to assist administrative decisions, not clinical ones. However, he pointed out that clinical freedom was restricted by resource constraints, which meant that we have to move beyond the medical ethic ‘to do good, and no harm’, to make decisions about where extra resources produced the greatest health benefi t (a social and citizen perspective).

AW agreed with me that both market failure and distributional goals consti-tute a need for CBA. But he had some qualifi cation regarding the distributional goals, and added a merit goods argument (‘assessment of need’) that was not in the thesis; i.e. ‘paternalism’ rather than ‘consumerism’. ‘Whether this is as it should be or not is a lively and controversial area of debate into which I will not enter at present. But if it is conceded that as a matter of fact it is quite a common phenomenon, then we have to ask whether the economist’s role is thereby emasculated (which I suspect is Jönsson’s position) or whether we should press on and try to systematise what will otherwise tend to be an


obscure area of unanalysed mystique (which is my preferred strategy, of which more later)’ (Williams, 1976).

While I advocated assessments of WTP for health improvements before illness strikes (insurance method) (Jönsson, 1977) to overcome the ability to pay objection, he was not totally convinced. ‘This [unequal income distribution in relation to need] can be overcome by experiments designed to elicit relative valuations in contexts in which purchasing power is constant, but it may be better to resort to “proxy” valuers, via the political process, or even to rely on “experts”, provided their values are brought into the open’. And he continues ‘. . . it may well be more effi cient to delegate the decision to people who know more about it, even if they are imperfect judges of consumers’ valuations’, and ‘. . . Even if they were “well-informed”, patients may not be the best judges of their own welfare’ (Williams, 1976).

I interpreted consumers here to be citizens/the general population. Perhaps that is a mistake, and AW instead meant ‘patients’. If that is the case, he is consistent with his later position not to use valuations of patients in the actual health states for valuation of the utility to be in those states.

AW ON THE WAY FORWARDAlternatives to Paretian Welfare Economics

There is an alternative view, which I espouse, which is that the economist’s con tribution to CBA is not restricted by the confi nes of Paretian Welfare Economics, but that, since we have special skills and expertise in handling valuation problems, we can (and should) go boldly into the territory of the valuation of ‘need’ by practitioners and policy makers as well as analysing com-pensation awards by judges and any other sources of information on ‘social’ values (Williams, 1976).

Systematic studies of practitioners and the concepts of health they use as well as the relative weights (valuations) they attach to the various dimensions of health. Work with medical researchers in the development of health status indexes (Williams, 1976).

Valuation of healthiness per se:

This work is not constrained within the Paretian straitjacket, but nor is it mystical and obscurantist and beyond analysis, for it has its own clear conceptual basis and rigorous measurement criteria (Williams, 1976).

And some advice to a PhD student:

I would hope to persuade thoughtful scholars like Bengt Jönsson that the pure Paretian doctrine, à la Mishan, has outlived its usefulness, and is now merely a


starting point for work in this fi eld, and one which we can safely move beyond in the fi eld of health economics (Williams, 1976).

IMPLICATIONS AND CONSEQUENCES OF HIS POSITIONAW understood early on that CBA is a valuable tool for improving resource allocation in health care, and he was very effective in communicating this insight. He stressed that CBA can and should be based on good science, and that CBA should ‘clarify’ the resource allocation problem, not make the decision. He did not take very great interest in discussing costs, but from what he wrote he took the position that CBA should be based on a social cost concept, and that distributional aspects should be addressed. The development of the QALY as an outcome measure was a major step forward in the practical application of economic evaluation in health care.

His critique of Paretian welfare economics as a theoretical base for CBA was very strong and effective, because it came from someone with deep knowledge about the subject. A consequence of this was that very few young health economists were brave enough to explore the opportunities of developing the WTP approach to valuation of health improvements. However, I interpret his objections based as much on practice as on principles. Welfare economics was used for formulating arguments against public health care, particularly in the US, and he was a fi rm defender of a public healthcare system. For a health economist coming from a country were there was unanimous support for an egalitarian publicly funded healthcare system, this was not a major issue. However, it is understandable if AW looked at it differently, since the NHS was founded in 1948 and was therefore less established than, for example, the public healthcare systems in the Nordic countries. Another important aspect, which AW probably sensed before other economists, was that WTP and the monetary valuation of health were rejected by decision makers, clinicians as well as administrators and politicians, in most healthcare systems in Europe, and probably also the US. My conclusion is that AW made a careful evaluation of what was needed for health economics to be accepted within the healthcare system. Even if WTP estimates could be adjusted to meet equity criteria, they may have been rejected by decisions makers.* For AW it was more important to improve practice than to be right in principle.

It also turned out that he advocated, and over 30 years developed, the right principles to be practical, making an outstanding contribution to making CBA a scientifi c method and an accepted and useful tool for improving health policy. AW did not need any ‘Paretian straitjacket’, but I think I share the conviction with BS that it is useful to have this available for CBA management consultants who do not have the integrity and intellectual standing of AW.

* Note the difference with CBA within environmental economics, where WTP is the gold standard, and where this method is much more developed.


Working closely with decision makers to help them to make decisions about the allocation of healthcare resources in the public interest is an important, but also a challenging, role. AW is a great example of how it can be done, and a role model for many of us who have tried.

REFERENCESCooper M and AJ Culyer (eds) (1973) Health Economics: selected readings.

Harmondsworth: Penguin.Gold MR, Siegel JE, Russell LB and Weinstein MC (1996) Cost-effectiveness in Health

and Medicine. New York: Oxford University Press.Jönsson B (1976) Cost–benefi t Analysis in Public Health and Medical Care [dissertation]

Lund University: Department of Economics.Jönsson B (1977) Application of cost benefi t analysis to health problems In: Halberstadt

V and Culyer AJ (eds) Public Economics and Human Resources. Proceeding of the 31st Congress of the International Institute of Public Finance. Paris: Editions Cujas, pp. 193–202.

Sugden R and Williams A (1978) The Principles of Practical Cost-benefi t Analysis. Oxford: Oxford University Press.

Wildavsky A (1973) If planning is everything, maybe it’s nothing. Policy Sciences 4: 127–53.

Williams A (1974a) Measuring the effectiveness of health care systems. British Journal of Preventive and Social Medicine 28: 196–202.

Williams A (1974b) Need as a demand concept (with special reference to health). In: Culyer AJ (ed.) Economic Policies and Social Goals. London: Martin Robertson, pp. 60–76.

Williams A (1974c) The cost|benefi t approach. British Medical Bulletin 30: 252–6.Williams A (1976) Cost-benefi t Analysis in Public Health and Medical Care – comments

on a thesis written by Bengt Jönsson. Lunds Universitet: Nationalekonomiska Institutionen. Meddelande 28 (mimeo, 13p). [Lund University: Department of Economics. Working paper 28 (mimeo 13p).]

Williams A (1977) Health service planning. In: Artis M and Nobay A (eds) Studies in Modern Economic Analysis: the proceedings of the Association of University Teachers of Economics, Edinburgh 1976. Oxford: Basil Blackwell & Mott Ltd, pp. 301–35.

Williams A (1981) Welfare Economics and Health Status Measurement. In: J van der Gaag J and Perlman M (eds) Health, Economics and Health Economics. Amsterdam: North Holland Publishing, pp. 159–70 and reprinted in: Culyer AJ and Maynard A (eds) (1997) Being Reasonable about the Economics of Health: selected essays by Alan Williams. Cheltenham: Edward Elgar Publishing Limited, pp. 107–19.

27

CHAPTER 4

The public–private challenge in health care

. . . Alan Maynard

INTRODUCTIONAlan Williams had a phenomenal capacity to teach and mentor, guiding generations of York and non-York colleagues along the paths of enlightenment, a process always based on his maxim: ‘Be reasonable! Do it my way!’. I was a grateful benefi ciary of his teaching and mentoring over the decades we worked together in York.

He was all too aware of the limitations of fi nance and delivery in health care, but robust in his defence of a healthcare system based on the principle of need, defi ned with typical precision as capacity to benefi t. At a time when that principle is under renewed challenge in healthcare systems across the world and as diverse as Australia, Canada, and Hong Kong, the purpose of this paper is to elaborate and extend the arguments that Williams deployed about confl icting ‘ideals’ and ‘actuals’ in the positions of those holding competing ideologies in the healthcare policy debate. The pertinence of this analysis is increased by the current confused policies of the British government, which, ever anxious to achieve greater effi ciency in the use of increased NHS funding, has adopted some innovations whose purposes appear to be inconsistent with the provision of care on the basis of need.

SOCIAL WELFARE FUNCTIONS AND COMPETING IDEOLOGIESOver 20 years ago, the Thatcher Government was challenging the UK’s National Health Service (NHS). The ‘social contract’ that had governed NHS politics since 1948 had broken down (Klein, 2001). Whereas previously there had been an unwritten concordat by which the medical profession was left to govern the delivery of health care which the government funded, the Thatcher


administration increasingly demanded to see evidence of ‘value for money’. This implied micromanagement of clinical activity and the doctors’ trade union, the British Medical Association, responded by demanding increased funding.

At the same time, and as part of the international and continuous cycle of healthcare reform, the Americans were debating the performance of their own healthcare system. At a conference in Washington DC in 1980, a York trio presented a paper including Williams’ description of the attitudes associated with differing ideological viewpoints (Culyer et al., 1981). Williams elaborated this subsequently with descriptions of the actual and idealised characteristics of the competing egalitarian and libertarian healthcare systems (Maynard, 1982; Maynard and Williams, 1984; Williams, 1988).

The three tables from these papers are shown next: attitudes associated with the two viewpoints (Table 4.1), the characteristics of an ideal healthcare system as advocated by adherents to the competing ideologies (Table 4.2), and the nature of the actual competing healthcare systems as critiqued by their opponents (Table 4.3). Adherents of each ideological position tend to criticise the actual performance (see Table 4.3) of their opponent with the idealised characteristics (see Table 4.2) of their own system, thereby fuelling rhetorical debate and avoiding addressing the resolution of well-established effi ciency and equity defi ciencies.

TABLE 4.1 ATTITUDES TYPICALLY ASSOCIATED WITH VIEWPOINTS A AND B

Viewpoint A (libertarian) Viewpoint B (egalitarian)

Personal responsibility

Personal responsibility for achievement is very important, but weakened if people are offered unearned rewards. Moreover, they weaken the motive force that assures economic well-being and thereby also undermine moral well-being, because of the intimate connection between moral well-being and the personal effort to achieve.

Personal incentives to achieve are desirable, but economic failure is not equated with moral depravity or social worthlessness.

Social concern

Social Darwinism dictates a seemingly cruel indifference to the fate of those who cannot make the grade.

Private charitable action is not rejected but potentially morally risky (because it may demean the recipient and corrupt the donor) and usually inequitable.

cont.

THE PUBLIC–PRIVATE CHALLENGE IN HEALTH CARE 29

Viewpoint A (libertarian) Viewpoint B (egalitarian)

Social concern (cont.)

A less extreme position is that charity, expressed and effected preferably under private auspices, is the proper vehicle, but it needs to be exercised under carefully prescribed conditions: for example, the potential recipient must fi rst mobilise his own resources and, when helped, must be in a less favourable position than the self-supporting (the principle of ‘lesser eligibility’).

It seems preferable to establish social mechanisms that create and sustain self-suffi ciency with precise rules of entitlement that are applied equitably and explicitly sanctioned by society at large.

Freedom Freedom is sought as a supreme good in itself. Compulsion attenuates both personal responsibility and individualistic and voluntary expressions of social concern. Centralised health planning and a large governmental role in healthcare fi nancing are an unwarranted abridgement of the freedom of both clients and health professionals; private medicine is therefore viewed as a bulwark against totalitarianism.

Freedom is the presence of real opportunities of choice; although economic constraints are less coercive than political constraints, they often represent the effective limits on choice. Freedom is not indivisible but may be sacrifi ced in one respect in order to obtain greater freedom in another. Government is not an external threat to individuals in society but is the means by which individuals achieve greater scope for action (that is, greater real freedom).

Equality Equality before the law is the key concept, with clear precedence being given to freedom over equality wherever the two confl ict.

Since the only moral justifi cation for using personal achievement as the basis for distributing rewards is that everyone has equal opportunities for such achievement, the main emphasis is on equality of opportunity; where this cannot be assured, the moral worth of achievement is undermined. Equality is the extension to the many of the freedom enjoyed by the few.

Source: This table was published in Williams A (1988) Priority setting in public and private health care: a guide through the ideological jungle. Journal of Health Economics 7: 173–83. Copyright Elsevier, 1988. Used and adapted with permission from Elsevier.


TABLE 4.2 IDEALISED HEALTHCARE SYSTEMS

Private PublicDemand 1 Individuals are the best judges

of their own welfare.1 When ill, individuals are frequently

imperfect judges of their own welfare.

2 Priorities determined by own willingness and ability to pay.

2 Priorities determined by social judgements about need.

3 Erratic and potentially catastrophic nature of demand mediated by private insurance.

3 Erratic and potentially catastrophic nature of demand made irrelevant by provision of free services.

4 Matters of equity dealt with elsewhere (e.g. in the tax and social security systems).

4 Since the distribution of income and wealth is unlikely to be equitable in relation to the need for health care, the system must be insulated from its infl uence.

Supply 1 Profi t is the proper and effective way to motivate suppliers to respond to demand (need).

1 Professional ethics and dedication to public service are the appropriate motivation, focusing on success in curing or caring.

2 Priorities determined by people’s willingness and ability to pay and by the costs of meeting their wishes at the margin.

2 Priorities determined by where the greatest improvements in caring or curing can be effected at the margin.

3 Suppliers have strong incentive to adopt least-cost methods of provision.

3 Predetermined limit on available resources generates a strong incentive for suppliers to adopt least-cost methods of provision.

Adjustment mechanism

1 Many competing suppliers ensure that offer prices are kept low and refl ect costs.

1 Central review of activities generates effi ciency; audit of service provision and management pressures keep the system cost-effective.

2 Well-informed consumers can seek out the most cost-effective form of treatment.

2 Well-informed clinicians prescribe the most cost-effective form of treatment for each patient.

3 If, at the price that clears the market, medical practice is profi table, more people will go into medicine and hence supply will be demand responsive.

3 If there is resulting pressure on some facilities or specialties, resources will be directed towards extending them.

cont.

THE PUBLIC–PRIVATE CHALLENGE IN HEALTH CARE 31

Private Public

Adjustment mechanism (cont.)

4 If, conversely, medical practice is unprofi table, people will leave it, or stop entering it, until the system returns to equilibrium.

4 Facilities or specialties on which pressure is slack will be slimmed down to release resources for other uses.

Success criteria

1 Consumers judge the system by their ability to achieve what they demand, when, where and how they want it.

1 Electorate judges the system by the extent to which it improves the health status of the population at large in relation to the resources allocated to it.

2 Producers judge the system by how good a living they make.

2 Producers judge the system by its ability to enable them to provide the treatments they believe to be cost-effective.

Source: This table was published in Williams A (1988) Priority setting in public and private health care: a guide through the ideological jungle. Journal of Health Economics 7: 173–83. Copyright Elsevier, 1988. Used and adapted with permission from Elsevier.

TABLE 4.3 ACTUAL HEALTHCARE SYSTEMS

Private PublicDemand 1 Doctors act as agents, mediating

demand on behalf of consumers.1 Doctors act as agents, identifying

need on behalf of patients.

2 Priorities determined by the reimbursement rules of insurance funds.

2 Priorities determined by the doctor’s professional situation, his assessment of the patient’s condition and the expected trouble-making proclivities of the patient.

3 Because private insurance coverage is a profi t-seeking activity, some risk rating is inevitable; hence, coverage is incomplete and uneven, distorting personal willingness and ability to pay.

3 Absence of direct fi nancial contributions at the point of service and of risk rating enables patients to seek treatment for trivial or inappropriate conditions.

4 Attempts to change the distri-bution of income and wealth independently are resisted as destroying incentives (e.g. the ability to buy better or more medical care if you are rich).

4 Attempts to correct inequities in the social and economic system by differential compensatory access to health services may encourage use of health care where it is unlikely to be cost-effective.

cont.


Private PublicSupply 1 What is most profi table to

suppliers may not refl ect consumer interests, and lack of clarity over consumers’ interests gives suppliers a range of discretion.

1 Personal professional dedication and public-spirited motivation may degenerate into cynicism if others are seen to be personally benefi ting from blatantly self-seeking behaviour.

2 Priorities determined by the extent to which consumers can be induced to part with their money and by the costs of satisfying the pattern of ‘demand’.

2 Priorities determined by what gives the greatest professional satisfaction.

3 Profi t motive strongly incentivises market segmentation and price discrimination and tie-in agreements with other professionals.

3 Since cost-effectiveness is not accepted as a proper medical responsibility, such pressures merely generate tension between the ‘professionals’ and the ‘managers’.

Adjustment mechanism

1 Professional, ethical rules are used to make overt competition diffi cult.

1 Few elaborate cost data required for billing purposes, so little useful information on costs routinely generated.

2 Consumers denied information about quality and competence and, since insured, may collude with doctors (against the insurance carriers) in infl ating costs.

2 Clinicians know little about costs, have no direct incentive to act on information they have, and sometimes have perverse incentives (i.e. cutting costs may make their life more diffi cult or less rewarding).

3 Entry into the profession made diffi cult and numbers restricted to maintain profi tability.

3 Very little known about the relative cost-effectiveness of different treatments and doctors wary of acting on any such information until a general professional consensus emerges.

4 If demand for services falls, doctors extend range of activities and push out neighbouring disciplines.

4 Phasing out of redundant facilities diffi cult because it often threatens the livelihood of some concentrated specialised group and has identifi able people dependent on it, whereas benefi ciaries are dispersed and only identifi ed as statistics.

cont.

the ideas and inﬂ uence of alan williams · 2020. 10. 20. · the ideas and inﬂ uence of alan...

Documents